Over the past several decades, the use of software in engineering has grown significantly. It is hard to find a field of engineering today that does not rely on one or more specialized engineering programs. Design engineers across all disciplines learn early on how to use the preferred program for their field. Often, out of frustration with the limitations of the software they are using, design engineers will often develop their own programs for very specific applications.
In 2018, it is impossible to get an engineering degree without learning at least a few of the more common programs like MATLAB or AutoCad. These software tools have become essential to today’s engineers and represent a significant portion of their education and training. Whether designing molecules, materials, parts or assemblies, software exists to expedite the process. Software programs have undeniably made the design process far more efficient.
One example where software has saved design engineers time and money is the electromagnetic compatibility (EMC) of systems within vehicles. As the number of antennas and circuits found in a car or a plane increases so too does the chance that they will interfere with each other, producing unexpected and sometimes dangerous results. It’s easy to take for granted that cars now come with Bluetooth, GPS, radar collision avoidance, tire pressure monitoring systems, and so forth. These technologies are increasingly designed to fit into smaller and smaller places.
Before EMC software, design engineers would have to individually create all the components to certain standards, and then bring the components into a test environment. Within this test environment, within close proximity of other devices, these components would be tested to see if they stayed within predefined parameters. If they did not, the components would have to be reconfigured until a satisfactory solution was achieved. This process was time-consuming, costly and limited innovation.
EMC software allows design engineers to simulate the test environment with software. The preliminary design-test-redesign cycle was replaced by software simulation until a fairly stable configuration is reached. Thus, when design engineers reach the point where the components are placed in the test environment, there is a high probability that the devices will work within designed parameters. The time and cost savings are such that more ambitious vehicle systems can be imagined and implemented.
Of course, with any powerful tool, software tools can be misunderstood or misused. Engineering software is only as reliable as its underlying assumptions. It is imperative that the engineer using the program fully understand its limitations and apply it only to appropriate problems. Otherwise, the engineer runs the risk of treating the device like a black box, simply trusting the results produced with no clear understanding of their reliability.
One example of this can be found in the world of material design. The field of computational chemistry started soon after the discovery of quantum mechanics in the early 20th century. At first the labor-intensive calculations were done by hand. One of the first applications of early computers was solving computational chemistry problems. The severe computational demand of even the most basic molecules, such as H2O, made any serious material design impractical. In an effort to expand the field's potential, in the 1970s a few assumptions were made to develop a new technique to greatly simplify calculations.
The assumptions made in this new technique were appropriate for calculating the energies of molecular systems containing atoms with lower atomic numbers. These assumptions made it possible to calculate the energies of much larger systems than in the past, which led to widespread adoption. However, a few decades later it was becoming apparent that the technique had severe limitations that its users were ignoring.
By the end of the 1990s, some scientists joined a movement to return to a pure first-principles technique for more accurate results. By then computational power had increased to the point where such an approach, while still computationally expensive, was possible for some systems. Today both approaches are used, but there are still those who ignore the original assumptions of the revised technique, treat the programs that implement the technique as black boxes, and publish results that are not meaningful. This practice confuses the field with unreliable results and slows progress.
Engineering software accelerates innovation. As AI becomes more widespread, it is being adopted in cooperation with engineering software to speed the design process up even more. The complexity of devices and systems has reached the point that designing and integrating them would be impossible without the help of software. As the complexity of these systems increases so too does the software to try and keep up. The danger is that eventually, the software will become so complex that fully understanding it will be beyond the engineer's ability. The process will then again become a black box, with sometimes unexpected results.
That is why it is essential that engineers stay on top of software developments and universities expand courses on software. In addition, soon engineers will have to have at least a basic understanding of AI. Although keeping up to date with software is tough, the innovations it makes possible justify the effort. There is no question, with the assistance of ever more effective engineering software, technological innovation will continue to accelerate across all engineering fields.